Rapidly after the onset of ischemia, cell respiration fails and the ATP pool is depleted. The cells depolarize, as they no longer can maintain ionic gradients across their membranes. In consequence, osmotically driven inflow of water into the cells leads to a reduction of the apparent diffusion coefficient (ADC) (12). ADC measurements, thus, show a significant reduction by approximately 30–40% in the ischemic core, depicting the ischemic territory very early and with sensitive contrast (13). Quantitative analysis of the degree of ADC reduction and correlation with the tissue area of energy breakdown has led to an ADC threshold for a severely damaged ischemic core region and a further threshold for the metabolic penumbra characterized by anaerobic glycolysis and tissue acidosis (Fig. 2) (14).

In a permanent occlusion model, the ADC reduction persists for a few days, then increases again to produce a pseudonormalization, followed by an ADC value above normal (15). These changes reflect the pronounced expression of the vasogenic edema during the first week post-stroke, followed by the cystic transformation. In an early reperfusion situation, the ADC may normalize very quickly in relation to rapid recovery of the energy metabolism (16, 17). With several hours delay, a secondary decline of the ADC occurs in the same territory that was characterized by ADC reduction at the end of the primary occlusion period. Thus, within the first few hours after successful reperfusion, the ADC may severely underestimate the damage in the ADC maps (Fig. 3).

Another parameter commonly applied is the T2 relaxation time. T2-weighted images show the early vasogenic edema as hyperintensity after a few hours (18). This hyperintensity persists and increases to its maximum at approximately 1 week after stroke onset. T2-weighted imaging, however, is not sensitive enough to detect the beginnings of T2 increases during the very early phase of the vasogenic edema with sufficient contrast (16). Quantitative T2 maps, on the other hand, allow the reliable detection of the ischemic territory even within an hour of stroke onset (16). This T2 increase is not reversible during early reperfusion when ADC transiently normalizes again. Moreover, the further development of T2 during this reperfusion phase – a further increase or stagnation – has been described as a good indicator for the severity of damage (16).

MR angiography provides information about the vascular occlusion and the position of a clot in thromboembolic stroke models (19). This contrast agent-free angiographic method allows repetitive screening for success of lysis. Together with perfusion MRI, it offers a detailed haemodynamic characterization of the occlusion period: vascular occlusion and reopening, tissue perfusion and collateral supply (despite occlusion of the main vessel).

T1-weighted MRI can be used together with iv injection of a gadolinium contrast agent to depict disturbances of the blood-brain barrier (BBB) (20). In the regions of open BBB, the contrast agent leaks out of the vascular bed into the parenchyma, inducing a T1 reduction in the tissue and thus resulting in a diagnostic hyperintensity.

In the chronic phase, ADC and T1-weighted imaging are of little help, particularly as the pseudonormalization phase may veil the pathophysiological processes, thus leading to false negative interpretation. T2-weighted imaging is the most widely used parameter for presentation of the lesion extent in the chronic phase. It should be noted, however, that even T2 maps carry the risk of false interpretation due to the development of T2 contrast over time: After the first wave of T2 increase during the pronounced expression of the vasogenic edema, T2 normalizes again in parallel with the resolution of the edema (21). This T2 normalization will continue if the lesioned tissue develops only selective neuronal death. In the case of pannecrosis with following cystic transformation, however, a second T2 increase occurs at this time. Thus, the time profile of T2 during the chronic phase is indicative of pannecrosis while a lack of T2 increase may be interpreted either as tissue recovery or selective neuronal death (21, 22).

Finally, during the late phase, several weeks after stroke induction, a hypointensity may appear in T2*-weighted images in some areas of the ischemic territory. The reason for this T2*-sensitive hypointensity was found to be iron accumulation in macrophages around vessels with delayed degradation. Iron ions of erythrocytes, entering the parenchyma through vascular leaks and incorporated by the macrophages, are the source of the image contrast, thus serving as sensitive monitor of delayed vessel degradation and ensuing macrophage activity (23, 24).

Manganese ions are taken up by cells in a similar way to calcium because of their same charge and van der Waals radius. After incorporation into vesicles, they are transported anterogradely along the axons and are even passed on transsynaptically (25, 26). Due to their electron configuration, manganese 2+ ions act as efficient T1 contrast agent – this effect is exploited for Manganese-Enhanced MRI, MEMRI. Thus, transport of manganese ions from the local injection site of manganese chloride can be followed in vivo. This can be used to trace connections and assess the intactness of a connection (27). Appearance of manganese-induced T1 contrast in the thalamus after local injection into the somatosensory cortex, thus, demonstrates the intact cortico-thalamic connection while lack of hyperintense signal in the thalamic nucleus would indicate loss of the axonal pathway.